Reactive & Functional Polymers 71 (2011) 782–790
Contents lists available at ScienceDirect
Reactive & Functional Polymers
journal homepage: www.elsevier.com/locate/react
Macroporous, responsive DNA cryogel beads
Nermin Orakdogen, Pinar Karacan, Oguz Okay ⇑
Istanbul Technical University, Department of Chemistry, 34469 Istanbul, Turkey
a r t i c l e
i n f o
Article history:
Received 27 January 2011
Received in revised form 15 April 2011
Accepted 23 April 2011
Available online 29 April 2011
Keywords:
DNA
Crosslinking
Macroporous
Swelling
Adsorbent
Carcinogens
a b s t r a c t
Biocompatible soft materials that are macroporous and tough are in demand for a range of applications.
Here, we describe the preparation of macroporous DNA cryogel beads by crosslinking DNA in frozen
aqueous solution droplets at 18 °C. Ethylene glycol diglycidyl ether was used as the crosslinker and
N,N,N0 ,N0 -tetramethylethylenediamine as catalyst. The beads swell in 4.0 mM NaBr 74–212 times their
dry weights and exhibit moduli of elasticity around 0.5 kPa. In dry state, they contain irregular large pores
of 101–102 lm in sizes due to the ice crystals acting as a template during the gelation reactions. DNA
beads can be compressed up to about 80% strain without any crack developments. They also exhibit
reversible swelling–deswelling cycles in water and acetone, respectively, undergoing a discrete phase
transition in aqueous acetone solutions containing 51% acetone. The ability of the beads for the removal
of carcinogenic agents from aqueous solutions was also demonstrated using phenanthrene as a model
compound. The sorption capacity of the beads was found to be 420 lg phenanthrene/g DNA.
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
All living cells contain deoxyribonucleic acid (DNA) molecules
carrying genetic information in their base sequences. In its native
form, DNA is a semi-flexible polymer with a double-helical (ds)
conformation stabilized by hydrogen bonds between the amine
bases [1]. When a DNA solution is subjected to high temperature,
the hydrogen bonds holding the two strands together break and
the double helix dissociates into two single flexible strands having
a random coil conformation [2]. Due to the unique structure of
DNA, chemical compounds having aromatic planner groups such
as benzopyrine and phenanthrene intercalate between adjacent
base pairs of ds-DNA and result in mutation and endocrine disruption [3,4]. This fact also suggests that, after insolubilization of DNA,
it can be utilized as an adsorbent for such toxic materials from
waters.
Several techniques were developed to prepare water-insoluble
DNA compounds for the removal of toxic materials. Yamada et al.
prepared water-insoluble and nuclease-resistant DNA films by
UV radiation [5], while Yamamoto and co-workers prepared DNA
liquid crystal gels by the reaction with metal cations [6,7]. DNA/
nanoparticle hybrids [8], DNA-immobilized nonwoven cellulose
fabric [9], DNA complexes with lipids [10], cationic surfactants
[11,12], DNA immobilized in gels [13,14], semi-interpenetrating
networks [15], DNA films [16], and hydrogels containing DNA
strands as grafts [17], or as crosslinks [18] are among materials
⇑ Corresponding author. Tel.: +90 212 2853156; fax: +90 212 2856386.
E-mail address: okayo@itu.edu.tr (O. Okay).
1381-5148/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.reactfunctpolym.2011.04.005
containing water-insoluble DNA. Although such materials can be
used as specific sorbent for the removal of carcinogens, they have
the general drawback that they are nonporous, and mechanically
weak. An efficient sorbent material should have a macroporous
structure allowing a fast liquid transport through the continuous
macropores [19]. In addition, a fast response against the external
stimuli and a good mechanical performance are also requirements
for efficient materials.
DNA hydrogel is a network of chemically crosslinked DNA
strands swollen in aqueous solutions [20]. Such soft materials are
a good candidate to make use of the characteristics of DNA such
as coil-globule transition, biocompatibility, selective binding, and
molecular recognition [16,18]. DNA hydrogels were prepared starting from branched DNA molecules via ligase-mediated reactions
[21]. These hydrogels can also be prepared by crosslinking DNA
in semi-dilute solutions using a chemical crosslinker such as ethylene glycol diglycidyl ether (EGDE) in the presence of N,N,N0 ,N0 tetramethylethylenediamine (TEMED) catalyst [22]. EGDE contains
epoxide groups on both ends that can react with the amino groups
on the nucleotide bases to form a three dimensional DNA network
[23]. DNA hydrogels are responsive systems exhibiting drastic volume changes in response to the external stimuli, such as the composition of aqueous solutions of acetone [22,24], or polyethylene
glycol [25], concentrations of inorganic salts [26–29], polyamines
[27,28], cationic macromolecules [30], or surfactants [27,28,31].
We have recently focused on developing responsive DNA hydrogels
with a wide range of tunable properties such as the conformation
of the network strands [25], viscoelasticity [32], and nonlinear
elasticity (strain hardening) [33]. Gelation reactions conducted at
N. Orakdogen et al. / Reactive & Functional Polymers 71 (2011) 782–790
50 °C show that a high concentration of DNA such as 9.3 w/v% stabilizes double-stranded (ds) DNA conformation while, at lower
concentrations, single stranded (ss) DNA gels were obtained [25].
No gel formation was observed at DNA concentrations as low as
5%. Gels formed from ss- or ds-DNA strands were highly transparent indicating that they are nonporous in dry state.
Preparation of DNA hydrogels with a macroporous structure
has, to our knowledge, not been previously reported. Creating an
interconnected pore structure within the crosslinked DNA network
would result in the formation of fast responsive DNA hydrogels. In
the present study, we describe the preparation of macroporous
DNA hydrogels in the form of millimeter-sized beads suitable as
specific adsorbent for carcinogenic compounds. Our strategy to
prepare such gel beads is to conduct the crosslinking reactions of
ds-DNA (about 2000 base pairs long) using EGDE crosslinker within the droplets of frozen DNA solutions at 18 °C. This low temperature gelation technique known as cryogelation is a simple route
for the preparation of macroporous gels [34–40]. During the freezing of an aqueous polymer solution containing a chemical crosslinker, the polymer chains and the crosslinker molecules expelled
from the ice concentrate within the channels between the ice crystals, so that the crosslinking reactions only take place in these unfrozen liquid channels. After crosslinking and, after thawing of ice,
a macroporous material (cryogel) is produced whose microstructure is a negative replica of the ice formed. As will be seen below,
the crosslinking DNA in aqueous frozen solution droplets at 18 °C
produces spherical, macroporous, tough cryogel particles with fastresponsivity. The characteristics of polyelectrolyte hydrogels such
as the volume phase transition can be observed in the cryogel
beads in a very short period of time and in a reversible manner.
Here, we also show that macroporous DNA beads can effectively
be used in the adsorption processes of carcinogenic compounds
such as dilute aqueous solutions of phenanthrene.
2. Experimental
2.1. Materials
Cryogels were made from deoxyribonucleic acid sodium salt
from salmon testes (DNA, Sigma). According to the manufacturer,
the% G–C content of the DNA used is 41.2%, and the melting temperature is reported to be 87.5 °C in 0.15 M sodium chloride plus
0.015 M sodium citrate. The molecular weight determined by
ultracentrifugation is 1.3 106 g/mol, which corresponds to
approximately 2000 base pairs. The crosslinker ethylene glycol
diglycidyl ether (EGDE, 50%, technical grade, Fluka), N,N,N0 ,N0 tetramethylethylenediamine (TEMED, Merck), phenanthrene (Fluka), and sodium bromide (NaBr, Merck) were used as received.
Stock solutions of EGDE and TEMED were prepared by dissolving
2.61 mL EGDE and 0.568 mL TEMED in 10 and 20 mL 4.0 mM NaBr,
respectively. DNA and TEMED concentrations in the gelation solutions were expressed as DNA% and TEMED%, respectively, which
are the mass of DNA and the volume of TEMED in 100 g reaction
solution. The crosslinker (EGDE) content of the reaction solution
was expressed as EGDE%, the mass of EGDE added per 100 g of
DNA.
2.2. Gelation reactions
The crosslinking reaction of DNA was carried out at 18 °C in
the presence of 50% EGDE. The pH of the reaction solution was
set to 11.0 by the addition of 0.44% TEMED [25]. DNA was first dissolved in 4.0 mM NaBr at 35 °C for 1 day. After addition of EGDE
and stirring for 1 h, TEMED was included into the reaction solution.
Note that the solutions containing more than 1% DNA were too
783
viscous and could not be dropped into the continuous phase to obtain spherical beads. Therefore, they were heated to 50 °C and held
at this temperature for 10 min to partially denature DNA and thus,
to decrease the viscosity of the solutions. Two techniques were
used for the preparation of macroporous DNA cryogel beads:
Technique A: DNA solution containing EGDE and TEMED was
added dropwise into an excess of liquid nitrogen to create small
frozen droplets. As each droplet touches the liquid nitrogen,
nitrogen starts to boil while the droplet spins on the liquid surface until it falls down into the liquid nitrogen. After obtaining
frozen droplets, they were transferred into paraffin oil as the
continuous phase at 18 °C and the reactions were conducted
for 3 days.
Technique B: DNA solution was directly dropped into paraffin oil
at 7 °C and after complete addition, the temperature of the oil
phase was decreased to 18 °C and the reactions were conducted for 3 days at this temperature. Note that droplets of
DNA solutions in paraffin oil cannot be obtained if the temperature was initially set to 18 °C due to the high viscosity of the
oil. Preliminary experiments conducted under different experimental conditions showed that 7 °C is the optimum temperature for the addition of the droplets into the oil phase.
After 3 days, the gel beads were removed from the oil phase,
washed several times with acetone. Thereafter, they were placed
in an excess of 4.0 mM NaBr solution and the solution was replaced
many times.
DNA cryogels were also prepared in the form of cylinders of
about 4.5 mm diameter. The reaction solution containing DNA,
EGDE, and TEMED prepared as described above was transferred
into plastic syringes. Half of the syringes were immediately frozen
in liquid nitrogen to prevent denaturation of DNA due to the effect
of EGDE-TEMED pair [25]. The other half of the syringes were
heated in an oven at 50 °C for 20 min to completely melt DNA, following quenching in liquid nitrogen to fix ss-DNA conformation.
Indeed, hyperchromicity measurements (see below) showed
90 ± 10% denaturation of ds-DNA in these samples All syringes
were then transferred into a freezer at 18 °C to conduct the gelation reactions for 3 days. In this way, cryogels consisting of mainly
ss-DNA and ds-DNA strands were prepared.
For comparison, hydrogels of DNA were also prepared by conducting the gelation reactions at 50 °C both in plastic syringes
and between the parallel plates of the rheometer (Gemini 150
Rheometer system, Bohlin Instruments) equipped with a Peltier
device for temperature control. The upper plate (diameter
40 mm) was set at a distance of 500 lm before the onset of the
reactions. During all rheological measurements, a solvent trap
was used to minimize the evaporation. Further, the outside of
the upper plate was covered with a thin layer of low-viscosity silicone oil to prevent evaporation of solvent. A frequency of x = 1 Hz
and a deformation amplitude c = 0.01 were selected to ensure that
the oscillatory deformation is within the linear regime.
2.3. Hyperchromicity measurements
For the hyperchromicity measurements, gelation reactions were
conducted, as described above, except that the frozen droplets or
the reaction solutions were transferred into empty glass vials at
18 °C instead of paraffin oil. After 5–10 min of the reaction time,
samples were taken and after thawing, they were diluted to a concentration of 26 mg/L with 4.0 mM NaBr. The degree of denaturation was estimated from the optical absorbance at 260 nm
measured with a T80 UV–visible spectrophotometer. The results
were presented as the normalized absorbance Arel with respect to
that measured from starting DNA solution. Because melting of
784
N. Orakdogen et al. / Reactive & Functional Polymers 71 (2011) 782–790
DNA strands leads to a rise of the normalized absorbance Arel up to
1.4 [41], the fraction of ss-DNA fragments in DNA (ss-DNA%) was
estimated as ss-DNA% = 250(Arel 1).
2.4. Swelling measurements
Swelling measurements were conducted on individual gel
beads equilibrium swollen in 4.0 mM NaBr (pH = 6.0). The beads
were placed separately in glass vials containing an excess of
4.0 mM NaBr at 21 °C. The swelling equilibrium was tested by
monitoring the diameter of the gel beads using an image analyzing
system consisting of a microscope (XSZ single Zoom microscope), a
CDD digital camera (TK 1381 EG) and a PC with the data analyzing
system Image-Pro Plus. The swelling equilibrium was also tested
by weighing the gel beads. Thereafter, the equilibrium swollen
gel beads were first added into acetone (poor solvent) and then
dried in vacuum at room temperature. The equilibrium volume
and the equilibrium weight swelling ratios of the beads, qv and
qw, respectively, were calculated as:
qv ¼ ðD=Ddry Þ3
ð1Þ
qw ¼ m=mdry
ð2Þ
where D and Ddry are the diameters of the equilibrium swollen and
dry beads, respectively, m and mdry are the weights of beads after
equilibrium swelling and after drying, respectively. Note that the
swelling measurements were conducted on at least six individual
beads prepared under the same experimental condition and the results were averaged.
For the measurement of the deswelling rates of gel beads, the
equilibrium swollen gel beads in NaBr solution were immersed
in acetone at 21 °C. The volume changes of beads were measured
in situ by following the diameter of the samples under microscope
using the image analyzing system. For the measurement of the
swelling rates of beads, the collapsed gel beads in acetone were
transferred into 4.0 mM NaBr at 21 °C. The diameter changes of
the beads were also monitored as described above. The results
were given as the relative volume swelling ratio Vrel = (Dt/D)3
where Dt is the gel diameter at time t. Swelling tests of the gels
in aqueous acetone solutions of various compositions were also
carried out as described above except that both the diameter and
the mass of the beads were monitored as a function of acetone
content.
For the determination of the swelling ratios of the cryogels in
the form of cylinders, the gels taken from the syringes were cut
into samples of about 5 mm in length and they were placed in an
excess of 4.0 mM aqueous NaBr solution. In order to reach swelling
equilibrium, the samples were immersed in solution for at least
4 weeks replacing the solution many times. After weighing the
samples, they were freeze-dried and the swelling ratio qw was calculated using Eq. (2). The gel fraction, that is, the mass of crosslinked DNA obtained from one gram of DNA was calculated from
the masses of dry, extracted DNA network and from the DNA mass
in the feed.
end-plate. The force F acting on the gel was calculated from the
reading of the balance m as F = mg, where g is the gravitational
acceleration. The resulting deformation DD was measured using
a digital comparator (IDC type Digimatic Indicator 543-262, Mitutoyo Co.), which was sensitive to displacements of 103 mm. The
force and the resulting deformation were recorded after 20 s of
relaxation. The measurements were conducted up to about 20%
compression. The weight loss of single beads during the measurement due to solvent evaporation or due to the applied force was
found to be negligible.
We have to mention that the interpretation of the compression
test data of spherical gel particles is complicated due to the significant variation of the contact area between the wall (PTFE endplate) and the originally spherical gel during deformation. For a
sphere with a constant volume during deformation, Hertz derived
the following equation for small deformation ranges [44–50].
F¼
4
GD0:5 DD1:5
3
ð3Þ
where G is the elastic modulus, and DD = D D0 , D and D0 are the
initial (swollen) undeformed and deformed diameters of gel sample,
respectively. The Hertz theory (Eq. (3)) considers the contact deformation of elastic spheres under normal loads in the absence of
adhesion and friction. According to Eq. (3), (3/4) F/D0.5 vs DD1.5 plots
should be linear for a given bead, with a slope equals to its elastic
modulus. Indeed, linear plots were obtained for all the beads studied in the range of deformation ratios above 5%. In Fig. 1, (3/4)FD0.5
is plotted against DD1.5 for swollen DNA gel beads formed at 3% and
5% DNA concentrations. The variations in the stress–strain curves
depending on the bead diameter were in the range of experimental
error.
2.6. Texture determination and porosity of beads
For the texture determination of dried gel beads, scanning
electron microscopy (SEM) studies were carried out at various
magnifications between 50 and 2000 times (Jeol JSM 6335F Field
Emission SEM). Prior to the measurements, network samples were
sputter-coated with gold for 3 min using Sputter-coater S150 B
Edwards instrument. The pore volume Vp of the gel beads was esti-
2.5. Elasticity tests
Uniaxial compression measurements were performed on individual gel beads in their swollen states. All the mechanical measurements were conducted in a thermostated room of 21 ± 0.5 °C.
The stress–strain isotherms were measured by using an apparatus
previously described [42,43]. Briefly, a swollen gel bead of 10–
14 mm in diameter was placed on a digital balance (Precisa 320
XB – 220A, readability and reproducibility = 0.1 mg). A load was
transmitted vertically to the gel through a rod fitted with a PTFE
Fig. 1. Typical stress–strain data of DNA gel beads as (3/4) F/D0.5 vs DD1.5 plots
according to Eq. (3). DNA concentrations and the technique used in the gel
preparation are indicated.
N. Orakdogen et al. / Reactive & Functional Polymers 71 (2011) 782–790
mated through uptake of methanol of the swollen gel beads. The
gel beads swollen in 4.0 mM NaCl were transferred into methanol
and methanol was refreshed several times until the mass of the
beads remains unchanged. Since methanol is a nonsolvent, it only
enters into the pores of the DNA network. Thus, Vp (mL pores in one
gram of dry DNA network) was estimated as
V p ¼ ðmM mdry Þ=ðdM mdry Þ
ð4Þ
where mM is the mass of the bead immersed in methanol and dM is
the density of methanol (0.792 g/mL).
2.7. Adsorption of carcinogens to DNA cryogel beads
For the determination of the adsorption capacity of DNA beads
for carcinogens, phenanthrene was used as a model polycyclic aromatic hydrocarbon. Taking into account its solubility range in
water (1.2 lg/mL), concentration of the test solutions was around
1 lg/mL. Stock solutions were prepared by dissolving 0.050 g of
phenanthrene in 50 mL acetone. Test solutions were prepared by
diluting 1 mL of the stock solution with water to 1L. Phenanthrene
removal tests were performed by placing 25–80 g of gel beads
swollen in 4.0 mM NaBr into 200 mL test solution and shaking
for 5 h at 21 °C and 170 rpm. Since phenanthrene is known to degrade through the process of photooxidation, the test solutions
were covered with dark sheets. Additionally, control solutions
were also prepared and checked for their phenanthrene contents
throughout the experiments. Samples were taken at certain time
intervals and the measurements of phenanthrene concentrations
were performed using fluorescence spectrometer (Perkin Elmer
LS 55) at excitation and emission wavelengths of 209 and
369 nm. FLWinlab software was used for calculations.
3. Results and discussion
Gelation reactions were carried out at 18 °C in frozen aqueous
solutions of DNA using EGDE as a crosslinker and TEMED as a catalyst. Previous work on DNA gelation at 50 °C showed that the gels
exhibit a maximum modulus of elasticity at pH = 11.0, corresponding to 0.44% TEMED [25,32]. This value of pH was fixed in the
present study. Preliminary experiments conducted at a DNA concentration of 3% showed that gel formation at 18 °C requires at
least of 50% EGDE (mass of EGDE per 100 g of DNA). Since the
molecular weights of EGDE and the nucleotide repeat unit of
DNA are 174.2 and 324.5 g/mol, respectively, and G-C content of
the DNA used is 41.2%, 50% EGDE corresponds to 9 mol of epoxide
groups added per mole of guanine base in the ds-DNA. This excess
785
amount of the epoxide groups required for gelation indicates existence of intramolecular crosslinking reactions. In the following, the
crosslinker content was set to 50% while the concentration of dsDNA in the reaction solution was varied between 1% and 5%.
At the usual crosslinking temperature of DNA, namely at 50 °C
[25,32], no gel formation was observed under these reaction conditions (0.44% TEMED and 50% EGDE). Rheological tests showed rapid decrease of the dynamic moduli of the reaction solution with
time, while hyperchromicity measurements indicated partial dissociation of the double helix into flexible single strand fragments.
For example, Fig. 2A–C shows the elastic modulus G0 (filled symbols) and the viscous modulus G00 (open symbols) during the crosslinking of ds-DNA at 50 °C. The solid and dashed horizontal lines
represent G0 and G00 of the initial DNA solutions before the addition
of EGDE and TEMED. Both moduli rapidly decrease and, at 3% and
5% DNA, they again increase at longer times. However, after 3 h of
reaction time, elastic moduli are below their initial values and, no
water-insoluble DNA gels were obtained. Fig. 2D shows the fraction of ss-DNA fragments (ss-DNA%) as a function of the reaction
time at a DNA concentration of 3%. Even at 25 °C, ss-DNA% rapidly
increases from 5% to 80% and, after heating to 50 °C; it becomes
90%. Further, almost complete melting of DNA was observed at
1% concentration while at 5%, the fraction of ss-DNA fragments increased to 63 ± 8%. Thus, instead of gelation, denaturation occurs
during the crosslinking of DNA at 50 °C and at or below 5% DNA
concentration.
However, DNA gels could be obtained by conducting the crosslinking reactions in frozen solutions at 18 °C. Two techniques
were used for the preparation of DNA gel beads: According to the
first technique (technique A), the aqueous solution containing
DNA, EGDE and TEMED was dropped into liquid nitrogen to create
small frozen organic droplets at 196 °C. Then, the frozen droplets
were transferred into paraffin oil at 18 °C as the continuous phase
and the reactions were carried out for 3 days. Images shown in
Fig. 3A–C were taken from frozen solution droplets in liquid nitrogen (A), from DNA beads just after preparation (B), and after equilibrium swelling in 4.0 mM NaBr. Freezing of DNA solution in
liquid nitrogen results in the formation of uniform frozen solution
droplets; the spherical shape of the frozen droplets remained unchanged after the crosslinking reactions as well as after swelling
of the crosslinked DNA particles in aqueous solutions. The occurrence of the gelation reactions below the freezing point of water
is due to the cryoconcentration of DNA in unfrozen domains of
the apparently frozen reaction system (Fig. 3D). It was shown recently that frozen hydrogel samples at temperatures between
10 and 24 °C contain 6% non-freezable water bound to the
polymer chains [51]. Thus, as water freezes, DNA strands, EGDE,
Fig. 2. (A–C) Elastic modulus G0 (filled symbols) and the viscous modulus G00 (open symbols) during the crosslinking of ds-DNA at 50 °C. EGDE = 50%. TEMED = 0.44%. DNA = 1
(A), 3 (B), and 5%. (C) The solid and dashed horizontal lines represent G0 and G00 of the initial DNA solutions before the addition of EGDE and TEMED. x = 1 Hz. c = 0.01. (D) The
fraction of ss-DNA fragments (ss-DNA%) shown as a function of the reaction time. The temperature is 25 and 50 °C before and after the dotted vertical line, respectively.
DNA = 3%. TEMED = 0.44%. EGDE = 50%.
786
N. Orakdogen et al. / Reactive & Functional Polymers 71 (2011) 782–790
(A)
(B)
(C)
(A)
(B)
(C)
(D)
(2)
(1)
DNA
EGDE + TEMED
DNA+ EGDE+TEMED
at high concentration
Ice crystals
Fig. 3. (A–C) Optical microscopy images of frozen DNA solution droplets in liquid nitrogen (A), after crosslinking reaction in paraffin oil at 18 °C (B), and after equilibrium
swelling in 4.0 mM NaBr (C). DNA = 3 (first row) and 5% (second row). EGDE = 50%. TEMED = 0.44%. All the scaling bars are 2 mm. (D) Scheme of a solution droplet before (1)
and after freezing at 18 °C (2).
Table 1
Properties of DNA gel beads formed at 18 °C. EGDE = 50%. TEMED = 0.44%. ssDNA% = the amount of ss-DNA fragments in DNA network chains. qw, qv = weight and
volume swelling ratios, respectively. Vp = total volume of the pores. G = modulus of
swollen beads. Standard deviations are indicated in parenthesis.
DNA%
Technique
ss-DNA%
qw
qv
Vp (mL/g)
G (kPa)
3
5
3
5
A
A
B
B
85
63
79
12
74 (14)
133 (10)
121 (20)
212 (45)
50 (15)
122 (9)
63 (17)
147 (27)
5.4 (1.7)
7.4 (2.9)
8.3 (0.9)
10 (2.5)
0.6
0.4
0.7
0.5
(10)
(8)
(10)
(10)
(0.1)
(0.1)
(0.2)
(0.1)
and TEMED are excluded from the ice structure and they accumulate in the unfrozen microregions, making the DNA concentration
in these regions high enough to conduct the crosslinking reactions
at 18 °C. The second technique for the preparation of DNA beads
(technique B) involves dropwise addition of the aqueous DNA solution containing EGDE and TEMED into paraffin oil as the continuous phase and crosslinking at 18 °C for 3 days. Similar beads as
shown in Fig. 3A–C were also obtained by technique B. Both techniques provide formation of DNA beads of diameters 2–3 mm in
dry or collapsed states, and 8–14 mm in swollen states. Note that
the size of the beads could be adjusted by changing the tip diameter (Dtip) of the pipettes from which the aqueous solution was
dropped into liquid nitrogen or into paraffin oil. In the present
work, Dtip was taken as constant at 2 mm.
Swelling, porosity, and elasticity characteristics of DNA beads as
well as the conformation of DNA network strands in terms of
Fig. 4. Equilibrium weight swelling ratio qw of ss-DNA (open symbols) and ds-DNA
cryogels (filled symbols) in the form of cylinders plotted against DNA concentration
at the gel preparation. EGDE = 50%. TEMED = 0.44%.
ss-DNA% are collected in Table 1. All the measurements were conducted on individual gel beads and the results were averaged. The
787
N. Orakdogen et al. / Reactive & Functional Polymers 71 (2011) 782–790
(A)
(B)
(A)
(B)
Fig. 5. SEM of DNA networks formed by technique A (upper images) and B (bottom images). DNA concentration at the gel preparation = 3 (A) and 5% (B). EGDE = 50%.
TEMED = 0.44%. The scaling bars are 10 lm. Magnification = 1000.
F/N
15
5 %, B
3 %, B
1
2
3
4
5 %, A
3 %, A
10
5
0
0.0
0.2
0.4
0.6
0.8
1.0
1- α
Fig. 6. Stress–strain data of DNA gel beads as the dependence of force F on the fractional deformation 1 a of the gel bead. Photographs show a swollen DNA gel bead formed
at 18 °C by technique B during the compression test. After compression, addition of 4.0 mM NaBr converts the gel bead back to its initial state. DNA concentration = 5%.
gel beads formed at 1% DNA were too weak to obtain reproducible
results. Due to the preheating period of the reaction solutions, DNA
network strands consist of 12–85% ss-DNA fragments. The fraction
of ss-DNA fragments decreases with increasing DNA concentration
due to the simultaneous increase of the concentration of DNA
counterions (Na+), which stabilize ds-DNA conformation, as observed before by the addition of salts to dilute DNA solutions
[52]. Indeed, the degree of denaturation of ds-DNA in water at
90 °C was found to be 100% and 18% at 1% and 9% DNA concentrations, respectively. Table 1 also shows that the beads swell in
4.0 mM NaBr 74–212 times their dry weights. Increasing DNA concentration at the gel preparation also increases their weight swelling ratios while the modulus of elasticity G of the swollen beads
remains around 0.5 kPa. We have to mention that the large
standard deviations in the measured values given in Table 1 are
due to the bead-to-bead variation of the gel properties.
Due to the different degree of denaturation of the DNA network
strands in the gel beads, a definitive conclusion regarding the effect
of the chain conformation and DNA concentration on the gel swelling cannot be drawn. Moreover, the gel fraction, that is, the amount
of crosslinked DNA obtained from one gram of DNA cannot be
determined. This is due to the presence of the continuous oil phase
during the formation of DNA beads, which hindered the determination of the bead mass at the state of gel preparation. Therefore,
experiments were repeated under the same experimental conditions but in plastic syringes at 18 °C. As described in the experimental part, ds-DNA and ss-DNA cryogels in the form of cylinders
were obtained. The gel fraction was found to be unity between 2%
788
N. Orakdogen et al. / Reactive & Functional Polymers 71 (2011) 782–790
180
Phenanthrene / μg
150
120
90
60
Fig. 7. The normalized volume Vrel of DNA gel beads shown as a function of the time
of deswelling in acetone and re-swelling in aqueous 4.0 mM NaBr. DNA = 3%.
Technique used: A (open symbols), and B (filled symbols). EGDE = 50%.
TEMED = 0.44%. The first and second deswelling–reswelling cycles are shown by
circles and triangles, respectively. The inset shows a semi-logarithmic plot of the
swelling data.
30
0
0
30
60
90
120
150
180
210
240
270
Time / min
mrel
Fig. 9. The amount of phenanthrene in the external solution shown as a function of
contact time with DNA gel beads formed by technique B. DNA = 5%. EGDE = 50%.
TEMED = 0.44%. 0.34 g beads were used for 200 mL of a phenanthrene solution of
initial concentration 0.835 lg/mL.
100
10-1
10-2
0
10
20
30
40
50
60
Acetone v/v %
70
80
90
100
Fig. 8. Variation of the swelling ratio mrel of DNA gel beads with the acetone
content of the external acetone/water mixture. Swelling temperature = 21 °C.
EGDE = 50%. TEMED = 0.44%. DNA = 3%, Method B (N). DNA = 5%, Method A (j)
and B (d). The dashed vertical line illustrates the phase transition region.
and 5% DNA indicating that DNA incorporates completely into the
network structure. Fig. 4 shows the equilibrium weight swelling
ratio qw of ds-DNA (filled symbols) and ss-DNA cryogels (open
symbols) plotted against DNA concentration. In contrast to the
large standard deviations in qw values of the cryogel beads (Table
1), most of the error bars in the figure are smaller than the symbols. In accord with the results obtained using cryogel beads, the
higher the DNA concentration, the larger is the swelling capacity
of the cryogels. Thus, the results suggest decreasing crosslinking
efficiency of EGDE as the concentration of DNA in the reaction
solution is increased. This unexpected behavior could be related
to diffusion restrictions of DNA strands in nonfrozen phase [53],
the extent of which increase with increasing DNA concentration.
Another interesting finding is that ss-DNA cryogels swell more
than ds-DNA cryogels indicating the effect of DNA conformation
on the gel swelling. Since ss-DNA is a flexible polymer with a persistence length p of about 1 nm, as compared to semiflexible dsDNA with p = 50 nm [54], the critical overlap concentration c of
ss-DNA is 80-fold larger than that of ds-DNA [25]. For example,
c of an aqueous solution of DNA of molecular weight of 2000 base
pairs (bp), as used in the present experiments, increases from 0.043
to 3.2 w/v% during denaturation [25]. Thus, for a given DNA concentration, ds-DNA strands are more overlapped than ss-DNA
strands in the reaction solution, which favors the occurrence of
intermolecular EGDE crosslinks and leads to the formation of gels
exhibiting a lesser degree of swelling.
Table 1 also shows that the total volume of the pores in the DNA
beads estimated from the uptake of methanol is 5–10 mL/g. To
visualize the pores in DNA gels, the network samples were investigated by scanning electron microscopy (SEM). Fig. 5 shows SEM
images of the samples prepared by techniques A (upper panel)
and B (bottom panel). DNA networks have a porous structure consisting of irregular large pores of sizes 101–102 lm. As schematically shown in Fig. 3D, the formation of a macroporous structure
in DNA gels is due to the presence of ice domains in the gelation
system. The crosslinking reactions of DNA only proceed in the unfrozen liquid channel around the ice crystals so that, after melting
of ice, a porous material is produced whose pore walls consist of
crosslinked DNA strands of high concentration.
The mechanical properties of DNA gels were investigated in
their swollen states in 4.0 mM NaBr by the compression tests. Linear stress–strain curves were obtained by application of Hertz
equation to the force–deformation data of the gel beads (Fig. 1).
An important point was that the beads can be compressed up to
about 80% strain without any crack development. This behavior
is illustrated in Fig. 6 where the force F acting on the gel bead is
plotted against the fractional deformation (diameter change/initial
diameter, 1 a). The mechanical stability of the beads increased
with increasing DNA concentration and, the beads formed by technique B were more stable than those formed by technique A. Photographs in Fig. 6 demonstrate how a DNA gel bead prepared at 5%
DNA according to technique B sustains a high compression. The
bead remains mechanically stable up to about 80% compression.
During the compression step and as the bead is squeezed under
the piston, it releases its water so that it can be compressed to high
strains. After the release of the load and, after addition of aqueous
NaBr, the bead immediately recovers its original shape, as seen
from the image denoted by 4 in Fig. 6.
N. Orakdogen et al. / Reactive & Functional Polymers 71 (2011) 782–790
Fig. 7 shows the response rate of DNA beads prepared at 3% DNA
concentration, where the normalized gel volume Vrel (volume of gel
at time t/equilibrium swollen volume in water) is plotted as a function of the time t of deswelling in acetone and re-swelling in water.
In order to check the durability of the gel beads against the volume
changes, this swelling-deswelling cycle was repeated twice. It is
seen that, in contrast to the very long equilibration time of the conventional DNA gels [30], the beads attain their equilibrium collapsed states within 10 min. Further, the re-swelling period of
the beads occurs much faster; they return back to their initial equilibrium swollen state within less than 1 min. All these cycles were
reversible and, no substantial differences were observed between
the beads prepared at various concentrations and by use of both
techniques. The fast responsivity of the gel beads is due to their
large pore volumes, which provide easy penetration of the solvent
molecules within the DNA network structure.
Since DNA is a polyelectrolyte, single DNA molecules as well as
the gels derived from DNA undergo volume phase transition in response to changes in the external stimuli, as typically observed in
synthetic ionic hydrogels [22]. The fast responsivity of the macroporous DNA gel beads offer the advantage that the phase transition
phenomena in DNA can be observed over a short period of time
and in a reversible manner. We investigated the phase transition
in DNA beads in water–acetone mixtures induced by a change in
solvent composition. For this purpose, the mass of a single DNA
bead was monitored as a function of the solvent composition. Each
bead which is originally in 4.0 mM NaBr was first immersed in an
excess solution containing 10 v/v% acetone and after 20 min, the
bead was transferred into the next concentrated solution. This procedure was conducted in up and down directions between 0% and
100% acetone and, could be completed within 1 day. Note that due
to the non-uniformity of the diameter of the beads in collapsed
state, the measurements were conducted by gravimetry. The results are presented in Fig. 8 where the relative mass mrel of the
bead is plotted against the acetone concentration. At low acetone
concentrations the beads are swollen, at high concentrations the
beads are collapsed. A discrete phase transition was observed at
51 ± 1% acetone, during which the bead mass changes about eight
times. In this concentration range marked with the vertical dashed
line in the figure, only swollen or collapsed states could be observed but never an intermediate state.
It is interesting to examine the abilities of DNA beads for the
adsorption of carcinogenic agents. As a demonstration, the beads
prepared at 5% DNA by technique B with 88% ds-DNA were immersed in 200 mL of an aqueous solution of phenanthrene of concentrations between 0.8 and 1 lg/mL, which is close to its
solubility limit (1.2 lg/mL). Since phenanthrene molecules are nonionic carcinogens, they are adsorbed only by intercalation into the
DNA double helices. After 2 h of contact time, the concentration
of phenanthrene could be reduced by 70%. Fig. 9 shows the amount
of phenanthrene in the external solution plotted against the contact
time with DNA beads. The amount of phenanthrene in the solution
rapidly decreases in the first 30 min but then approaches a limiting
value. The maximum sorption capacity of the beads after 4 h was
found to be 420 lg phenanthrene/g dry bead with 10% standard
deviation. It is seen that, even in such dilute solutions, phenanthrene can be effectively removed using the DNA beads. Experiments conducted using DNA gels in the form of rods consisting of
100% ds-DNA gave the same sorption capacity for phenanthrene.
4. Conclusions
We described the preparation of macroporous DNA hydrogel
beads by crosslinking DNA using EGDE as a crosslinker in frozen
aqueous solution droplets at 18 °C. DNA beads of diameters
789
2–3 mm in dry or collapsed states, and 8–14 mm in swollen state
were obtained at a DNA concentration between 3% and 5%. The
beads swell in 4.0 mM NaBr 74–212 times their dry weights and
exhibit moduli of elasticity around 0.5 kPa. In dry state, they contain irregular large pores of 101–102 lm in sizes due to the ice crystals acting as a template during the gelation reactions. DNA beads
can be compressed up to about 80% strain without any crack
developments. They also exhibit completely reversible swelling–
deswelling cycles in water and acetone, respectively, i.e., they return to their original volume after a short period of time. The gel
beads deswell in aqueous acetone solutions of increasing acetone
content undergoing a discrete phase transition in solutions containing 51% acetone. The ability of the beads for the removal of carcinogenic agents from aqueous solutions was also demonstrated
using phenanthrene as a model compound. The sorption capacity
of the beads was found to be 420 lg phenanthrene/g DNA.
Acknowledgments
Work was supported by the Scientific and Technical Research
Council of Turkey (TUBITAK). O.O. thanks Turkish Academy of Sciences (TUBA) for the partial support.
References
[1] J. Xu, T.H. LaBean, S.L. Craig, in: A. Ciferri (Ed.), Supramolecular Polymers, CRC
Press, 2004, p. 445 (Chapter 12).
[2] V.A. Bloomfield, D.M. Crothers, I. Tinoco Jr., Nucleic Acids: Structures,
Properties, and Functions, University Science Books, Sausalito, 2000.
[3] L.S. Lerman, J. Mol. Biol. 3 (1961) 18.
[4] P.J. Dandliker, R.E. Holmin, J.K. Barton, Science 262 (1993) 1025.
[5] M. Yamada, K. Kato, M. Nomizu, K. Ohkawa, H. Yamamoto, N. Nishi, Environ.
Sci. Technol. 36 (2002) 949.
[6] T. Dobashi, K. Furusawa, E. Kita, Y. Minamisawa, T. Yamamoto, Langmuir 23
(2007) 1303.
[7] K. Furusawa, M. Wakamatsu, T. Dobashi, T. Yamamoto, Langmuir 23 (2007)
10087.
[8] R.C. Mucic, J.J. Storhoff, C.A. Mirkin, R.L. Letsinger, J. Am. Chem. Soc. 120 (1998)
12674.
[9] M. Yamada, K. Kato, M. Nomizu, M. Haruki, K. Ohkawa, H. Yamamoto, N. Norio,
Biomaterials 22 (2001) 3121.
[10] K. Tanaka, Y. Okahata, J. Am. Chem. Soc. 118 (1996) 10679.
[11] M.C. Moran, M.G. Miguel, B. Lindman, Biomacromolecules 8 (2007) 3886.
[12] M.C. Moran, M.G. Miguel, B. Lindman, Langmuir 23 (2007) 6478.
[13] K. Iwata, T. Sawadaishi, S. Nishimura, S. Tokura, N. Nishi, Int. J. Biol. Macromol.
18 (1996) 149.
[14] H. Kitamura, E. Matsuura, A. Nagata, S. Sakairi, S. Tokura, N. Nishi, Int. J. Biol.
Macromol. 20 (1997) 75.
[15] W.G. Liu, X.W. Li, G.X. Ye, S.J. Sun, D. Zhu, K.D. Yao, Polym. Int. 53 (2004) 675.
[16] N. Ishizuka, Y. Hashimoto, Y. Matsuo, K. Ijiro, Colloids Surf., A 284–285 (2006)
440.
[17] Y. Murakami, M. Maeda, Macromolecules 38 (2005) 1535.
[18] Y. Murakami, M. Maeda, Biomacromolecules 6 (2005) 2927.
[19] O. Okay, Prog. Polym. Sci. 25 (2000) 711.
[20] F.E. Alemdaroglu, A. Hermann, Org. Biomol. Chem. 5 (2007) 1311.
[21] S.H. Um, J.B. Lee, N. Park, S.Y. Kwon, C.C. Umbach, Nat. Mater. 5 (2006) 797.
[22] T. Amiya, T. Tanaka, Macromolecules 20 (1987) 1162.
[23] X. Lu, Y. Xu, C. Zheng, G. Zhang, Z. Su, J. Chem. Technol. Biotechnol. 81 (2006)
767.
[24] N. Yonekura, H. Mutoh, Y. Miyagi, E. Takushi, Chem. Lett. 954 (2000).
[25] F. Topuz, O. Okay, Biomacromolecules 10 (2009) 2652.
[26] F. Horkay, P.J. Basser, Biomacromolecules 5 (2004) 232.
[27] D. Costa, M.G. Miguel, B. Lindman, J. Phys. Chem. B 111 (2007) 8444.
[28] D. Costa, M.G. Miguel, B. Lindman, J. Phys. Chem. B 111 (2007) 10886.
[29] H. Mayama, T. Nakai, E. Takushi, K. Tsujii, K. Yoshikawa, J. Chem. Phys. 127
(2007) 034901.
[30] D. Costa, M.G. Miguel, B. Lindman, Adv. Colloid Interface Sci. 158 (2010) 21.
[31] D. Costa, P. Hansson, S. Schneider, M.G. Miguel, B. Lindman,
Biomacromolecules 7 (2006) 1090.
[32] F. Topuz, O. Okay, Macromolecules 41 (2008) 8847.
[33] N. Orakdogen, B. Erman, O. Okay, Macromolecules 43 (2010) 1530.
[34] V.I. Lozinsky, S.A. Morozowa, E.S. Vainerman, E.F. Titova, M.I. Shtil’man, E.M.
Belavtseva, S.V. Rogozhin, Acta Polym. 40 (1989) 8.
[35] V.I. Lozinsky, Russ. Chem. Rev. 71 (2002) 489.
[36] F. Plieva, X. Huiting, I.Yu. Galaev, B. Bergenstahl, B. Mattiasson, J. Mater. Chem.
16 (2006) 4065.
[37] H. Kirsebom, B. Mattiasson, I.Yu. Galaev, Langmuir 25 (2009) 8426.
[38] D. Ceylan, O. Okay, Macromolecules 40 (2007) 8742.
790
N. Orakdogen et al. / Reactive & Functional Polymers 71 (2011) 782–790
[39] D.C. Tuncaboylu, O. Okay, Eur. Polym. J. 45 (2009) 2033.
[40] D.C. Tuncaboylu, O. Okay, Langmuir 26 (2010) 7574.
[41] K.L. Planken, G.H. Koenderink, R. Roozendaal, A.P. Philipse, J. Colloid Interface
Sci. 291 (2005) 120.
[42] N. Gundogan, D. Melekaslan, O. Okay, Macromolecules 35 (2002) 5616.
[43] F. Topuz, O. Okay, React. Funct. Polym. 69 (2009) 273.
[44] H. Hertz, Über die Berührung Fester Elastischer Körper (on the Contact of
Elastic Solids). J. Reine und Angewandte Mathematik 92 (1881) 156; In: D.E.
Jones, G.A. Schott (Eds.) Miscellaneous Papers b H. Hertz. London, Macmillan,
1896.
[45] A. Knaebel, S.R. Rebre, F. Lequeux, Polym. Gels Networks 5 (1997) 107.
[46] B.J. Briscoe, K.K. Liu, D.R. Williams, J. Colloid Interface Sci. 200 (1998) 256.
[47] M. Barquins, M.E.R. Shanahan, Int. J. Adhes. Adhes. 17 (1997) 313.
[48] W.-M. Lu, K.-L. Tung, S.-M. Hung, J.-S. Shiau, K.-J. Hwang, Powder Technol. 116
(2001) 1.
[49] M.N. Inci, B. Erman, O. Okay, S. Durmaz, Polymer 42 (2001) 3771.
[50] D. Melekaslan, N. Gundogan, O. Okay, Polym. Bull. 50 (2003) 287.
[51] M.V. Dinu, M.M. Ozmen, E.S. Dragan, O. Okay, Polymer 48 (2007) 195.
[52] C. Schildkraut, S. Lifson, Biopolymers 3 (1965) 195.
[53] H. Kirsebom, G. Rata, D. Topgaard, B. Mattiasson, I.Yu Galaev, Macromolecules
42 (2009) 5208.
[54] P.J. Hagerman, Annu. Rev. Biophys. Biophys. Chem. 17 (1988) 265.